Lingfeng Shen

My name is Lingfeng Shen(沈凌峰), a research scientist at Bytedance (San Jose). I'm currently working on the foundation models and their applications. Before that, I obtained my M.S. degree at Department of Computer Science, Johns Hopkins University, and I received my B.S. degree (Mathematics) from China Agricultural University.

During my time at Johns Hopkins, I wokred with Daniel Khashabi and Philipp Koehn. Besides, I was fortunate to spend my last year of undergraduate at Tencent AI Lab, mentored by Haiyun Jiang, Lemao Liu and Shuming Shi.

My current research interests include:

I used to be interested in adversarial robustness [Link], [Link], backdoor robustness [Link], and evaluation [Link], [Link]

Email: starkshen3333[at]gmail.com; lingfeng.shen[at]bytedance.com

Github   /   Google Scholar   /   Semantic Scholar   /   Twitter   /   LinkedIn     

I use a so-called "geometric owl" as my avatar in most places.

The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts
Lingfeng Shen, Weiting Tan, Sihao Chen, Yunmo Chen, Jingyu Zhang, Haoran Xu, Boyuan Zheng, Philipp Koehn, Daniel Khashabi
ACL 2024  
paper  /  code  

Do pretrained Transformers Really Learn In-context by Gradient Descent?
Lingfeng Shen*, Aayush Mishra*, Daniel Khashabi
ICML 2024 (Oral)  
paper  /  code  

The Trickle-down Impact of Reward (In-)consistency on RLHF
Lingfeng Shen,  Sihao Chen, Linfeng Song, Lifeng Jin, Baolin Peng, Haitao Mi, Daniel Khashabi, Dong Yu
ICLR 2024  
paper  /  code  

On the Evaluation Metrics for Paraphrase Generation
Lingfeng Shen, Lemao Liu, Haiyun Jiang, Shuming Shi
EMNLP 2022 (Oral)  
paper  /  code